Adaptive neuro fuzzy controller for adaptive compliant robotic gripper
Highlights
► Controlling input displacement of a new adaptive compliant gripper. ► This design of the gripper with embedded sensors as part of its structure. ► A new and original principle for adaptive grasping. ► The handling of irregular, unpredictably shaped and sensitive objects.
Introduction
Grasping is one of the most frequent subjects to deal with in robotics owing to the requirement of moving or manipulating different objects. In spite of the various benefits of the grasping techniques developed to realize grasping processes, their current limitations make them expensive and with low flexibility. The key point in the grasping system is the gripper. The gripper performance is very important when fragile objects of different stiffness and shapes are manipulated and hence a reliable force control is crucial. This problem can be overcome with the use of deformable or flexible fingers which improve the limited capabilities of robotic rigid fingers (Becedas et al., 2011, Petković and Pavlović, 2012).
The requirement for flexible fingers is the ability to make safety grasps, detect and recognize objects (Chitta et al., 2010, Droessler et al., 2001, Russell and Wijaya, 2003, Staretu and Itu, 2011). Their modalities of applications differ highly from conventional grippers since conventional grippers are equipped in a domestic environment and are usually not intended for repetitive tasks that require high precision or strong force. For such purposes, when conventional grippers must be equipped with sensors, vision sensors are popular choices (Abdullah et al., 2012, Allen and Bajcsy, 1987, Bohg and Kragic, 2010, Jimenez et al., 1997, Radig and Florczyk, 2001). However, they may have limitations in such environmental conditions as in darkness, in very dirty or dusty situations, in foggy conditions, or even underwater. Being equipped with embedded sensors (Issa, Petković, Pavlović, & Zentner, 2011) is a good choice because the information of object properties is directly provided with the influence of the environmental condition. Embedded sensors also offer great potential for improving the grasp synthesis in object recognition and manipulation due to their good sensitivity and capability of detection and recognition of grasping object. The embedding sensing capability allows changing the gripper manipulation strategy in real time to achieve an adaptive grasp. To improve the control of a robotic gripper, fuzzy logic (FL) or artificial neural network (ANN) control has attracted much attention in recent years.
The concept of fuzzy theory was first introduced by Zadeh, 1965a, Zadeh, 1965b and is used to describe dynamic systems that are too complex and/or too ill-defined to synthesize controllers using conventional mathematical modeling techniques. Mamdani (1974) applied the fuzzy set theory for developing fuzzy logic controllers (FLCs) for controlling dynamic systems, and since then many more researchers have developed FLCs for various applications.
ANNs are a family of intelligent algorithms which can be used for time series prediction, classification, and control and identification purposes. Neural networks have an ability to train with various parameters. As a non-linear function, they can be used for identifying the extremely nonlinear system parameters with high accuracy. Neural networks can learn from data. However, understanding the knowledge learned by neural networks has been difficult. In contrast, fuzzy rule based models are easy to understand because they use linguistic terms and the structure of IF-THEN rules. Unlike neural networks, however, fuzzy logic by itself cannot learn. Since neural networks can learn, it is natural to merge these two techniques. This merged technique of the learning power of the ANNs with the knowledge representation of FL has created a new hybrid technique, called neuro fuzzy networks or adaptive neuro fuzzy inference system (ANFIS) Jang, 1993. ANFIS, as a hybrid intelligent system that enhances the ability to automatically learn and adapt, was used by researchers for modeling (Ghandoor and Samhouri, 2009, Nguyen and Ngo, 2008, Singh et al., 2012, Yetilmezsoy et al., 2011), predictions (Ghandoor and Samhouri, 2009, Hosoz et al., 2011, Khajeh et al., 2009, Sivakumar and Balu, 2010) and control (Altin et al., 2012, Areed et al., 2010, Kurnaz et al., 2010, Ravi et al., 2011, Tian and Collins, 2005) in various engineering systems. The basic idea behind these neuro-adaptive learning techniques is to provide a method for the fuzzy modeling procedure to learn information about data (Aldair and Wang, 2011, Dastranj et al., 2011). The ANFIS is one of the methods to organize the fuzzy inference system with given input/output data pairs (Grigorie and Botez, 2009, Wahida Banu et al., 2011). This technique gives fuzzy logic the capability to adapt the membership function parameters that best allow the associated fuzzy inference system to track the given input/output data (Akcayol, 2004, Moustakidis et al., 2008, Omar et al., 2011, Peymanfar et al., 2010).
In this paper, the application of ANFIS is proposed to control the input displacement for the gripper. A control algorithm based on embedded sensors voltage changing was derived to perform tasks of detection and recognition of a grasping object. Simultaneously, the controller provided the input displacement signal according to the shape of the fixed grasping object. The fingers of the robot gripper were equipped with embedded sensors of conductive silicone rubber (Petković and Pavlović, 2011, Valenta, 2008). To evaluate the shape recognition algorithm, many experiments of object-grasp touching exploration were conducted with the robotic gripper. The gripper should recognize an object within the first 5 mm of the input displacement.
Section snippets
Adaptive flexible gripper with embedded sensors
The development of universal grippers able to pick up unfamiliar objects of widely varying shapes and surfaces is a very challenging task. Passively compliant underactuated mechanisms are one way to obtain the gripper which could accommodate any irregular and sensitive grasping objects. The purpose of the underactuation is to use the power of one actuator to drive open and close motion of the gripper. The underactuation can morph shapes of the gripper to accommodate different objects. As a
ANFIS controller design
A controller is a device which controls each and every operation in a decision-making system. From the control system point of view, it brings stability to the system when there is a disturbance, thus safeguarding the equipment from further damage. It may be a hardware-based controller or a software-based controller or a combination of both. In this section, the development of the control strategy for control of the gripper input displacement is presented using the concepts of ANFIS control
Development of Simulink model
The Simulink model for the control of input displacement of the gripper was developed in Matlab (Fig. 9). This Simulink model with the ANFIS controller was developed using the various toolboxes available in the Simulink library. The entire system modeled in Simulink was an opened loop control system consisting of controllers, samplers, comparators, constants, the mux, gain blocks, constant blocks, ANFIS editor blocks, output sinks (scopes), input sources, etc. Here, it can be seen that the
Simulation results
After many experimental simulations, the input displacement of gripper and change in voltage of the embedded sensors were obtained. These results were separated in the three clusters which represented a particular grasping pattern. Each grasping pattern required a different input displacement so the results were stored with 2 inputs and one output. The ANFIS editor was opened in the command window (Fig. 10a). These variables, which were in the form of data in the workspace, were loaded into the
Conclusion
Robotic grasping of unknown objects remains an open problem in the robotic community. The handling of irregular, unpredictably shaped and sensitive objects introduces demands on gripper flexibility and dexterity. Reaching the dexterity and adaptation capabilities requires the control of a lot of actuators and sensors. The dexterity can also be obtained by underactuation, which consists in equipping the finger with fewer actuators than the number of degrees of freedom. In manipulating objects by
References (38)
- et al.
dSPACE based adaptive neuro-fuzzy controller of grid interactive inverter
Energy Conversion and Management
(2012) - et al.
Adaptive neuro-fuzzy control of an induction motor
Ain Shams Engineering Journal
(2010) - et al.
Learning grasping points with shape context
Robotics and Autonomous Systems
(2010) - et al.
An adaptive neuro-fuzzy inference system model for predicting the performance of a refrigeration system with a cooling tower
Expert Systems with Applications
(2011) - et al.
Featureless classification of tactile contacts in a gripper using neural networks
Sensors and Actuators
(1997) - et al.
Application of adaptive neuro-fuzzy inference system for solubility prediction of carbon dioxide in polymers
Expert Systems with Applications
(2009) - et al.
Adaptive neuro-fuzzy inference system based autonomous flight control of unmanned air vehicles
Expert Systems with Applications
(2010) - et al.
An adaptive neuro-fuzzy tracking control for multi-input nonlinear dynamic systems
Automatica
(2008) - et al.
Design adaptive neuro-fuzzy speed controller for an electro-mechanical system
Ain Shams Engineering Journal
(2011) - et al.
Design of a general proposed neuro-fuzzy controller by using modified adaptive-network-based fuzzy inference system
International Journal of Electronic Commerce
(2010)
Estimation of elastic constant of rocks using an ANFIS approach
Applied Soft Computing
Adaptive neuro-fuzzy control of a flexible manipulator
Mechatronics
An adaptive neuro-fuzzy approach for modeling of water-in-oil emulsion formation
Colloids and Surfaces A: Physicochemical and Engineering Aspects
Fuzzy sets
Information and Control
Object exploration using a three-axis tactile sensing information
Journal of Computer Science
Application of adaptive neuro-fuzzy controller for SRM
Advances in Engineering Software
Design an intelligent controller for full vehicle nonlinear active suspension systems
International Journal on Smart Sensing and Intelligent Systems
Cited by (133)
Design of type-2 fuzzy logic controller in a smart home energy management system with a combination of renewable energy and an electric vehicle
2023, Journal of Building EngineeringAdaptive neuro fuzzy selection of the most important factors for photovoltaic pumping system performance prediction
2020, Journal of Building EngineeringFuzzy Logic-Controlled Gripper Force Feedback for Haptic Device
2024, Journal of Advanced Computational Intelligence and Intelligent InformaticsVariable Stiffness Soft Robotic Fingers Using Snap-Fit Kinematic Reconfiguration
2023, IEEE Transactions on Robotics